CtrlK
BlogDocsLog inGet started
Tessl Logo

the-provider-runtime-symmetry

Keep Stave's Claude and Codex provider runtimes in sync when editing one of them. Use when a change touches `electron/providers/claude-sdk-runtime.ts`, `electron/providers/codex-app-server-runtime.ts`, executable lookup, env construction, CLI session launch env, tooling status probes, or any provider-agnostic behavior. Trigger on phrases like "provider runtime", "claude adapter", "codex adapter", "sibling adapter", "PATH env", "CLAUDE_CONFIG_DIR", "CODEX_HOME", "tooling status", "어댑터 대칭", "양쪽 프로바이더 확인".

74

Quality

Does it follow best practices?

Impact

No eval scenarios have been run

SecuritybySnyk

Passed

No known issues

SKILL.md
Quality
Evals
Security

Quality

Content

85%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is highly actionable and well-sequenced, with concrete file paths, a numbered symmetry checklist, and explicit verification feedback loops. Its only weakness is mild redundancy between the checklist and the "Common Failure Modes" section, which keeps conciseness just below a clean 3.

Suggestions

Consolidate the 'forked env' / option-leak cases so each failure mode appears once — either in the Symmetry Checklist or in 'Common Failure Modes' — with cross-references rather than restated detail, to tighten conciseness.

Consider trimming the 'Required Check Files' table to files not already named inline in the checklist, or vice versa, to reduce repeated path mentions.

The 'Runtime options' checklist line is very dense (it lists six wiring targets in one bullet); splitting it into sub-bullets would preserve actionability while improving scanability.

DimensionReasoningScore

Conciseness

The body is dense and assumes Claude's competence (no basic-concept padding), but several items repeat across sections — e.g. the forked-env issue appears in both checklist §4 and "Common Failure Modes" #2, and several file paths are restated in both the "Required Check Files" table and inline. It is mostly efficient but could be tightened, so it sits at score 2 rather than a clean 3.

2 / 3

Actionability

Guidance is concrete and executable: exact files ("electron/providers/cli-path-env.ts", "electron/main/ipc/schemas.ts"), specific grep targets ("electron/providers/, src/store/app.store.ts"), and a real command ("bun run typecheck"), matching the copy-paste-ready anchor; for an instruction-only skill this fully actionable guidance is not penalized for lacking code.

3 / 3

Workflow Clarity

The "Symmetry Checklist" gives a numbered 6-step sequence and "Verification" supplies explicit checkpoints (typecheck always; start a turn with Claude; start a turn with Codex) plus a feedback loop for probe-vs-runtime drift ("Do not verify env changes with probes only"), matching the explicit-validation-and-checklists anchor.

3 / 3

Progressive Disclosure

There are no bundle files; the SKILL.md is self-contained with well-organized sections and only one-level-deep, clearly signaled references to sibling skills and docs/providers/codex-upgrade-checklist.md. Per the rubric's simple-skill note and the bundle-structure guideline, well-organized sections with no deep nesting earn a 3.

3 / 3

Total

11

/

12

Passed

Description

100%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is specific, well-triggered, and complete: it states a concrete action, gives explicit 'Use when' guidance, and lists natural trigger phrases (including bilingual variants). It uses third-person voice throughout with no over-claims.

DimensionReasoningScore

Specificity

The description states a concrete action ("Keep Stave's Claude and Codex provider runtimes in sync") and enumerates specific trigger surfaces ("executable lookup, env construction, CLI session launch env, tooling status probes"), matching the multiple-specific-actions anchor rather than the single-domain score-2 example.

3 / 3

Completeness

It answers "what" (keep runtimes in sync) and "when" via an explicit "Use when a change touches..." clause plus a "Trigger on phrases like..." list, satisfying the explicit-triggers anchor; it is not missing or only-implied as at score 2.

3 / 3

Trigger Term Quality

It lists natural terms a developer would actually say ("provider runtime", "claude adapter", "sibling adapter", "PATH env", "CLAUDE_CONFIG_DIR", "CODEX_HOME", "tooling status") plus Korean variants, giving broad coverage rather than the partial score-2 set.

3 / 3

Distinctiveness Conflict Risk

It is tightly scoped to Stave's Claude/Codex provider-runtime parity with distinct file paths and trigger phrases, forming a clear niche unlikely to fire for unrelated skills; it does not read generically like the score-1/2 examples.

3 / 3

Total

12

/

12

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository
sendbird/stave
Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.